A novel Road User Safety Field Theory for traffic safety assessment applying video analytics
DOI: 10.5204/thesis.eprints.234039
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This research addresses the limitations of conventional road safety assessments, which rely on reactive, police-reported crash data that suffers from under-reporting and limited behavioral detail. While traffic conflict techniques offer a proactive alternative, existing methods often ignore application-specific contexts, fail to incorporate crash severity, and overlook heterogeneities in traffic environments and driver behavior. To resolve these gaps, the study introduces the Road User Safety Field (RUSF) theory, a novel framework that integrates crash frequency and severity estimation into a unified safety assessment model. The methodology combines extensive systematic literature reviews with empirical data analysis. The author reviewed 549 studies for concept mapping and 386 recent studies to identify prevalent conflict metrics and thresholds. Primary data was collected via video surveys at four signalized intersections in Southeast Queensland over four weekdays. A deep learning-based computer vision system automatically extracted road user trajectories, including speeds, distances, and accelerations, as well as conflict metrics such as Time-to-Collision (TTC), Modified Time-to-Collision (MTTC), Deceleration Rate to Avoid a Crash (DRAC), and Proportion of Stopping Distance (PSD). Delta-V, representing the expected post-collision velocity change, was used as a crash severity indicator. The study employed multivariate extreme value theory and copula theory to model the joint distributions of these conflict indicators and Delta-V, estimating crash frequency by severity. Additionally, the transferability of these models was tested across different intersections. The RUSF theory conceptualizes safety as a "buffer" around road users, generating a repulsive "Risk Force" that increases as users approach each other, peaking at collision. This force accounts for momentum changes, thereby predicting outcome severity. Results from the rear-end crash case study demonstrated that the Risk Force model significantly outperformed traditional conflict indicator-based models in predicting both severe and non-severe crashes. The study identified MTTC, DRAC, and Delta-V as the optimal indicator combination for rear-end crash risk in organized traffic environments. Furthermore, the developed crash frequency-by-severity models proved transferable to similar sites with minimal calibration. The significance of this work lies in providing a comprehensive, adaptable method for traffic safety assessment that eliminates the need for context-dependent indicator selection. By explicitly modeling the entire crash mechanism from normal interaction to collision, the RUSF theory enhances the accuracy of road user movement models, such as car-following and lane-change simulations. This framework has implications for long-term road design, real-time applications like adaptive signal control, and the integration of Connected and Automated Vehicles (CAVs) by enabling the modeling of crash risks for both equipped and unequipped road users.
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-06-20 |
| archive | success | canonical_url | — | — | 1 | 2026-06-26 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-20 |
| chunk | success | chunk | — | — | 1 | 2026-06-20 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-20 |
| promote | success | — | — | — | 1 | 2026-06-20 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-20 |
| verify | success | — | — | — | 1 | 2026-06-26 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Empirical Findings: crash risk outcomes